UTHOPHTSES IN THE OBSIDIAN OF THE EOCCHE EOSSE. 443 



blown up by the liberated steam. Iddings holds that the separa- 

 tion of anhydrous silicates and silica releases the water in the 

 region of crystallization. Approaching the problem with the 

 advantage of the considerations put forward by these writers, we 

 venture to affirm that the steam-vesicles have developed first, and 

 independently of any crystallizing tendency ; and that only under 

 special conditions, such as prevail in the obsidians of Lipari, do 

 vesicles play a part in modifying their surroundings and assisting 

 spherulitic growth. Such modifications have been described by 

 Mr. Eutley ^ as arising during the artificial heating of an obsidian 

 from Montana, white crystalline pellets being formed in the interior 

 of the vesicles. 



The Eocche Eosse rock proves that Weiss ^ was right in stating 

 that spherulitic aggregation may take place around a vesicle ; and 

 Prof. Judd ^ has remarked, when describing the rocks of Lipari, 

 that in some cases the formation of spherulites " has been deter- 

 mined by the liberation, in the midst of the vitreous mass, of an 

 infinitesimal bubble of volatile matter." But it seems that this 

 structure is limited to highly silicated, viscid, and slowly moving 

 vitreous lavas, where individual steam-bubbles can exercise 

 prolonged hydrothermal action on the glass. We may mention in 

 conclusion that Delesse,^ in the almost unrivalled plates illustrating 

 his ' Eecherches sur les Eoches globuleuses,' has figured an obsidian 

 from Ischia, which has clearly the same type of lithophysul 

 structure as that which we have endeavoured to describe. 



APPENDIX. 



On LiTHOPHTSEs and Hollow Spheeulites in alteeed Eoce:s. 

 By Prof. Geenville A. J. Cole, P.G.S. 



Prom the date of the publication of Mr. Iddings's memoir on 

 Obsidian Cliff, above referred to, it has been felt, I think, that Von 

 Eichthofen's view of the primary origin of the cavities of lithophyses 

 had received very valuable support. It became clear, at any rate, 

 that, from one cause or another, spherulites might be formed with 

 more or less regular hollows in their interior. It wi]l now be 

 necessary, in dealing with any old example, to endeavour to distin- 

 guish between primary and secondary hollows ; and I fear that in 

 some cases the distinction must remain a matter of opinion. The 

 Eocche Eosse obsidian shows that we may expect the material of 

 a ' spherulite ' some 5 mm. or more in diameter to have a composite 

 and many-centred character when it has accumulated within and 

 round a vesicle. Traces of the ' bridges ' may also be left, as I 

 thiuk is the case in some of the Ordovician examples so ably dealt 



^ ' On Alteration induced by Heat in certain Vitreous Eocks,' Proc. Eoy. 

 Soc. vol. xl. (1886) p. 435. 



2 Zeitschr. d. Deiitsch. geol. Gesellsch, vol. xxix. (1877) p. 421, 



^ ' Contributions to the Study of Volcanos,' Geol. Mag. for 1875, p. 65. 



* Mem. Soc. geol. de France, 2me ser. tome iv. pi. xxv. fig. 11 ; and p. 361. 



